8 citations · 17 across the 24 of their papers we have counts for
20 papers · 1 filter
Big-Big-SycoBench: A Benchmark to Detect Sycophancy Through Model Responses to Valid vs. Invalid Arguments
Marthe Ballon, Arne Vanhoyweghen, Seorin Kim +11
Sycophancy in large language models risks amplifying unsupported claims or beliefs across the knowledge ecosystem as these systems increasingly contribute to research synthesis, ev…
A proof of the Karpelevič theorem via invariant polygons
Brecht Verbeken, Brecht F. Verbeken, Vincent Ginis
The Karpelevič theorem describes the complex numbers that occur as eigenvalues of stochastic matrices of a fixed order. We give a self-contained proof based on invariant polygons.…
Stochastic realisers of non-degenerate full-degree Type II reduced Ito polynomials
Brecht Verbeken, Brecht F. Verbeken, Vincent Ginis
Let , where , , , , and . These are the non-degenerate full-degree Type II reduced Ito polynomials…
The Eigenvalue Region of a Family of Markov Chains with Randomness Confined to Two States
Brecht Verbeken, Brecht F. Verbeken, Vincent Ginis
We determine the complete eigenvalue region of the following Markov chains on states, for . The states are and . Each transitio…
Shorter Reasoning, Earlier Answers? An Evaluation of Reasoning Interfaces
Francesca Carlon, Vincent Ginis, Andres Algaba
Large language models often reason at length before answering, increasing cost and latency. Prompts and trained settings can shorten this reasoning, but a shorter trace may only sh…
How Much Does a Reasoning Summary Reveal? An Observability Ladder for Large Language Models
Andres Algaba, Francesca Carlon, Lynn Delcon +3
Large language models often show users a final response and a short reasoning summary while the full reasoning trace stays hidden. We introduce an observability ladder that holds e…
Geometric Configurations of Perturbed Jailbreak Prompts
Lynn Delcon, Andres Algaba, Vincent Ginis
Perturbation techniques that turn unsuccessful jailbreak prompts into successful ones are continuously evolving, constituting a major security threat to LLM safety. In this paper,…
The Type III realisation conjecture of Kirkland and Šmigoc
Brecht Verbeken, Brecht F. Verbeken, Vincent Ginis
Kirkland and Šmigoc constructed a family of stochastic matrices realising the Type III boundary polynomials in the Karpelevič region and conjectured that, conversely, every stochas…
Neural Networks for Inverse Design of Cascaded-Mode Near-Field Landscapes
Wannes Luts De Martelaere, Joeri Lenaerts, Vincent Ginis
Structuring optical near-fields is important for applications in microscopy and nanoparticle manipulation. Traditionally, near-fields are structured using antenna nanostructures th…
Rebuttals Move Peer-Review Scores, but Initial-Review Structure Bounds the Movement
Mathieu Louis, Tibo Vanleke, Vincent Ginis +1
Author rebuttals are the main post-submission window in peer review, but their effect on reviewer scores remains hard to measure because score updates mix rebuttal content with ini…
Thinking Like a Scientist? A Structural Study of LLM-Generated Research Methods
Francesca Carlon, Brecht Verbeken, Brecht F. Verbeken +2
Large Language Models (LLMs) are increasingly used to guide research methodology, yet their default methodological tendencies under minimal prompting remain unclear. Here, we promp…
On the Spectral Region of n-Cycle Stochastic Matrices
Brecht Verbeken, Brecht F. Verbeken, Vincent Ginis
For every , we determine the union of the spectra of row-stochastic matrices supported on a directed -cycle with self-loops and strictly positive cycle edges. These matr…
Neutrino Fingerprints: Image-Based Encodings of IceCube Events for CNN Direction Reconstruction
Floriano Tori, Brecht Verbeken, Brecht F. Verbeken +1
Reconstructing the direction of incoming neutrinos in the IceCube Neutrino Observatory is an important problem in astrophysics. The public IceCube--Neutrinos in Deep Ice Kaggle com…
One in Eight OpenAlex Abstracts Has Integrity Issues
Seorin Kim, Vincent Holst, Vincent Ginis
Scientific abstracts are increasingly used as primary data in computational metascience research, yet the quality of these abstracts in widely used bibliographic databases has not…
Scalable Classification of Course Information Sheets Using Large Language Models: A Reusable Institutional Method for Academic Quality Assurance
Brecht F. Verbeken, Brecht Verbeken, Joke Van den Broeck +5
Purpose: Higher education institutions face increasing pressure to audit course designs for generative AI (GenAI) integration. This paper presents an end-to-end method for using la…
Probing Graph Neural Network Activation Patterns Through Graph Topology
Floriano Tori, Lorenzo Bini, Marco Sorbi +2
Curvature notions on graphs provide a theoretical description of graph topology, highlighting bottlenecks and denser connected regions. Artifacts of the message passing paradigm in…
Probing the Trajectories of Reasoning Traces in Large Language Models
Marthe Ballon, Brecht Verbeken, Brecht F. Verbeken +2
Large language models (LLMs) increasingly solve difficult problems by producing "reasoning traces" before emitting a final response. However, it remains unclear how accuracy and de…
Structurally Human, Semantically Biased: Detecting LLM-Generated References with Embeddings and GNNs
Melika Mobini, Vincent Holst, Floriano Tori +2
Large language models are increasingly used to curate bibliographies, raising the question: are their reference lists distinguishable from human ones? We build paired citation grap…
Benchmarks Saturate When The Model Gets Smarter Than The Judge
Marthe Ballon, Andres Algaba, Brecht Verbeken +2
Benchmarks are important tools to track progress in the development of Large Language Models (LLMs), yet inaccuracies in datasets and evaluation methods consistently undermine thei…
Turning Citation Networks Inside Out: Studying Science Using Content-Based Knowledge Graphs from LLM-Derived Taxonomies
Seorin Kim, Vincent Holst, Vincent Ginis
Scientific fields are often mapped using citations and metadata, despite knowledge being transmitted primarily through content. We introduce an 'inside-out' approach that reconstru…